Reimagine GenAI: Creative Applications Beyond Efficiency

Drew: Hello, Renegade Marketers! If this is your first time listening, welcome, and if you’re a regular listener, welcome back. Before I present this episode, I’m thrilled to announce the first-ever in-person CMO Super Huddle that we’re hosting in Palo Alto on November 8, 2024. The theme is “Daring Greatness in 2025” and we’re rocking a full slate of inspiring speakers with ample time for networking. Tickets are on sale now, so grab yours at cmohuddles.com. It’s gonna be flocking amazing!

You’re about to hear a Bonus Huddle where experts share their insights into the topics of critical importance to our B2B CMO community, CMO Huddles. The expert in this Huddle is Jenny Nicholson, founder of Queen of Swords, amazing human who shared some of the most creative use cases of Gen AI, I guarantee you’ll ever see. This episode should inspire you to think beyond using Gen AI for efficiency and just for content, say like blogs and so forth. It’s a tool that can transform the way B2B brands interact with employees and customers in just imaginative and important ways. And you’ll see it, you’ll hear it. You’ll go, “Oh my god, we got to do some of that.” If you like what you hear, please subscribe to the podcast and leave a review. You’ll be supporting our quest to be the number one B2B marketing podcast. Okay, let’s dive in.

Narrator: Welcome to Renegade Marketers Unite, possibly the best weekly podcast for CMOs and everyone else looking for innovative ways to transform their brand, drive demand, and just plain cut through. Proving that B2B does not mean boring to business. Here’s your host and Chief Marketing Renegade, Drew Neisser.

Drew: Hello, huddlers. I am excited to introduce you to Jenny Nicholson for a conversation on creative uses of Gen AI. I hope to blow your mind. Jenny describes herself as a creative thinker with too many ideas, too few hours in the day, and too much pride to ask for help. I saw her speak at the BRXND AI conference in New York in May and was absolutely blown away by her creativity and her ability to take these tools like LLMs and turn them into some amazing stuff. So hello, Jenny. How are you and where are you this wonderful day?

Jenny: I’m doing well. I’m at home in Durham, North Carolina. I just got back from four days and nights in the mountains of North Carolina in my travel trailer with no cell service or electricity. So it’s nice to be back to my computer and to the world.

Drew: Amazing. You were off the grid in the Blue Ridge, or…?

Jenny: Exactly, exactly.

Drew: Nice. Well, I’m a big fan of Durham, North Carolina, as you and I may have talked about. It is a wonderful city, and there are some, you know, places to study there too. Okay, so just in case the audience has to leave early, and we need to persuade them to stick around, can you offer two to three things many marketers get wrong when using large language models like, say, ChatGPT?

Jenny: I would say sort of with generative AI, in general, the thing that I see most companies, including marketers, but in general, getting wrong is there’s this sort of jump to efficiency, to using it as a way to save time, save money, cut down on having to use humans. And I think that’s wrong. I think the actual power of these, the promise of these, potential of these, is an expansive one. I do a lot of consulting with companies, and 70% of what people ask me for when it comes to generative AI, I’m like, that’s not an AI challenge. That’s an automation challenge. And while automation has some really important benefits, and there are some places where it can really help with efficiency, that’s not really the promise of this technology.

I always say that instead of looking at large language models or generative AI as a way to do what you’ve always done, but faster and cheaper and with fewer people, what would happen if you used it to actually do things that you couldn’t have done before? And that’s always been sort of my focus, and what I wish more companies would do. You know, Toys R Us just yesterday or the day before came out with the world’s first commercial that was generated with Sora, which is OpenAI’s text-to-video model. I don’t know if anybody saw it, but it wasn’t a good ad. There was not an interesting idea, and it was—apologies if anybody here’s from Toys R Us, happy to talk with you about what I would have done differently. But you know, it was just kind of a bummer, because it didn’t move anything forward. It just sort of was an okay ad that even without Gen AI would have been just an okay ad.

Drew: You really hit on the fundamental reason why we are here today, which is to open folks’ minds to the kinds of things that you can create. I agree with you that a lot of what I see is about, oh, well, you can create something that took four hours, might now take two hours, and that sounds very appealing but that’s not necessarily something that is going to lead to better content, more creative content, more engaging content, more engaging programs. And that really is where we get to sort of creativity. So I’m okay if we want to jump from there, as in, we’re not talking about efficiency. We’re really talking about creative applications and creating things that you probably couldn’t have done before. And so maybe we start with show and tell even earlier than expected.

So I want to make sure that we frame this so everyone in the audience or almost everyone in the audience, they’re B2B marketers, and there’s an overwhelming sort of sense of rational and features, and we have to, you know, give them the value proposition in black and white, and you’ll get three times the ROI in three months and all of that. So this is going to be hard for them because a lot of your examples are B2C. But what I want folks to try to do is step back for a second and say, Jenny solved problems. Saw challenges and solved problems. Solve them with some tools. And let’s just start because this is so fun with one who created, like all of them at the kitchen table, with “Any Day Is a Party” or “Turn Any Day Into a Party” and talk about this one.

Jenny: Before I get into that, though, I want to make something really clear. This idea of sort of B2B having to be a certain way, and B2C getting to be a different way, something that I’ve never sort of agreed with and never kind of followed that rule, even before I did anything with generative AI. As I told you, Drew, one of the first things I started doing in advertising that kind of moved me away from the sort of standard formula was I started making games. And one of the games that I made was for an enterprise voice and data networking solutions company that was a 50-level game that the only way to get from one level to the next was to find the password. And people thought that nobody would do it. And it was incredibly successful, both with the IT decision-makers we were trying to connect with, but it also went viral with teenagers in the Philippines, and they all—like you can even find now, people reminiscing, “Remember that summer we were all obsessed with Ultimate Problem Solver?” And so I think it’s important just to remember that no matter who your audience is, whatever title they have, whatever company they work at, at the end of the day, they’re still people, and the things that work on people work on them too.

Anyways, now that my little speech is over, I’ll talk about “Turn Any Day Into a Party.” So I have a 10-year-old, and you know, sometimes you have long weekends where you don’t know what to do, and it’s really tempting to put them in front of the television. And this is a really handy way to get out of that. It’s a custom GPT that turns any day into a party. So you go, you say, “I want a party today,” it searches the Internet to figure out whatever weird holiday is happening. So I don’t know if you guys know, but today is National Chocolate Pudding Day. Congratulations. And so the party that Instant Party came up with for today is the Pudding Paradise Party. It gives you decor, food ideas, drink ideas, costume ideas, activity ideas. So you see here there’s going to be a chocolate pudding eating contest, DIY pudding bar, chocolate trivia, and a movie screening of Willy Wonka and the Chocolate Factory. And then there’s also a playlist, and you’ll see that they’re all chocolate and sugar and candy-themed because of the party. And then it also even makes for you an invitation that has all the signatures of ridiculous AI-generated type, which, in my mind, actually just makes it more delightful.

Drew: By the way, so two key things is, I want this day, you know, the playlist with “Sugar, Sugar” by the Archies resonates with me. And I’m sure many of my generation are—part one. And just reminding me to watch Charlie in the Chocolate Factory, the question is, which version did they recommend?

Jenny: You know, it did not specify. I mean, I’m a big believer that the Gene Wilder version is the only version that counts, though the new, new one with Timothée Chalamet was actually surprisingly good.

Drew: So that’s right, there’s three so. And have you done any of these days with your daughter?

Jenny: I have. Yeah.

Drew: And how has she reacted?

Jenny: It’s fun because it’s just sort of like on a random Saturday, like at 11, I’m like, “Want to have a party,” and she’s like, “Yes.” They somehow always tend to involve chocolate, though. I don’t know why our days have always ended up on chocolate. There was a version I did before this one just to—because I was playing with it this morning. Because, like I said, today is also like National Beauticians Day, and there was a Glam Jam Party. But this one was a little more fun to me.

Drew: The reason why I wanted to start here is a target audience for CMOs are there teams, who are often remote, and this feels like something, a game that you could play remotely with a team, also thinking about just overall employee engagement and how hard it is for companies to bring employees together. How long did it take you to actually create this little application?

Jenny: The whole thing, to get it to work and everything? A couple of hours.

Drew: And so what about the Spotify integration? How did you actually pull this—

Jenny: This one doesn’t actually have a Spotify integration. This one just the playlist.

Drew: So it just figures it out. It goes out in the world and finds all the songs with chocolate in it, or candy or sugar. So using a tool to create sort of an entertainment idea. Does it automatically find out what the day is, or you have to tell it what the day is?

Jenny: No, I just… there are two buttons. When you first come to the GPT, there’s a button that says, “I want to party today,” and it uses today’s date, and then you can say, “I want to party later.” And then it’ll ask you what date you have in mind.

Drew: And then it takes it from there. Okay, that is really fun. Okay, so hopefully you’re now thinking about, folks in the audience are thinking about, huh, this is interesting. Hadn’t thought about using this as an employee engagement tool or team building tool. That’s fun. Okay, let’s move on.

Jenny: Okay, so this is, you know, one of the things that I really like is, I like learning. I’m probably the most curious person I know. And so one of the things that I really love about large language models is any curiosity I have, they can match it, and they know all kinds of things about all kinds of things that I don’t know. And so this is a curiosity engine where you just start by either saying, “I’m curious about this, tell me about it,” or “Tell me something surprising.” If you ask it to tell you something surprising, nine times out of 10, it’s going to tell you about octopi, because, honestly, octopi are freaking crazy. But anyways, it gives you some information. Tells you a surprising thing, but then gives you sort of three more paths you can wander down. So in this case, it told me about octopi, then asked me if I wanted to learn about other animals with superpowers, like the immortal jellyfish that can theoretically live forever, to learn about extremophiles, alien life forms on Earth, or to get into the science of blood across different animals in the animal kingdom, from the green blood of reptiles to the clear blood of ice fish. And so I really love this sort of idea, that I can start anywhere I want and then kind of wander through learning, and I think a lot about what this might be like for education, what this might be like for schools, what this might be like for teaching people, instead of sort of passively sitting back, kind of getting entertainment delivered to them. What would it be like if there was a world where they could sort of follow their own curiosity and see where it led.

Drew: What was this like to create? What did you actually have to build in order for it to work?

Jenny: I said you are an endless curiosity engine. Somebody asks you for this, you give them this, then you give them three more paths they can go down, and then you do that forever until they’re done.

Drew: So it’s a GPT, by the way on octopi. Was it My Octopus Friend, the movie?

Jenny: My Octopus Teacher.

Drew: My Octopus Teacher, oh, my God, my wife cannot eat octopi.

Jenny: Yeah, yeah. That was, that was an interesting one. It’s very interesting to see the difference in the sort of reception on that one between men and women. That’s another, another topic, but I think that it’s a really interesting point, like, essentially, I just told it what I wanted it to do.

Drew: I mean, that’s insight, right? You have to tell it what to do, and you have to give it parameters, and that is part of the educational process. But you’ve spent a lot of time with these tools, and so you have a sense of what you have to tell it, right? I mean, you’ve worked with it long enough that you know, and I think that’s one of the hard parts for folks just getting started, is they don’t actually know how to sort of use it even.

Jenny: I think what’s really interesting, especially if we’re talking about sort of the base foundation models, GPT, Gemini from Google, and Anthropic’s Claude, a lot of working with these models is actually, and this is why I get a little bit nervous about these sort of like, put in campaign, put in brand, put in challenge, hit button, get campaign. A lot of the work of working with large language models is actually the work of learning how big the gap is between what you think you’re saying and what you’re actually saying. So a lot of the work that I’ve been doing over the almost three years I’ve been working with these large language models is work on refining my own communication skills and really realizing that, like, “Oh, this is what I thought I was saying. This is what I was actually saying.” Because when you talk to people and they don’t have any idea what you’re talking about, what do they do? They smile, they nod, they say, “Oh, of course, totally.” A large language model just tries to do exactly what you ask for. They’re trained to be helpful assistants. So they’re trained to whatever you ask for, give you exactly what you ask for. And so often people… People will ask for something, they’ll get something that’s not what they had in their mind, and then they’re like, “Well, this tool sucks.” When it’s actually a matter of like, well, actually, maybe it does, but often you find that you’re not communicating really clearly. And so I think it’s very funny that there’s this, you know, world of prompt engineering and all of this stuff. And, you know, chain of thought and tree of thought and this and that and the other thing. But when it comes down to it, prompt engineering is really knowing what to say, so that you say what you actually mean and not what you think you’re saying.

Drew: Well, there’s part of that. And then also just how the machine responds, one of the good exercises that I that I love. So, you know, I have this series of Penguin images that I use on posts on Saturdays, my sort of, my rants, of weekly rants. And I have an ongoing prompt for DALL-E for it. And what I love about it is, when I put a, let’s do a version of this, in this situation, it shows me an image, and then it tells me what it heard. And it’s really interesting. It’s not what I said, but it gives me something back. And I go, “Okay, that’s interesting.” So there is this disconnect, right? I think it’s helpful that the machine actually tells me the parameters that it used to create the images.

Jenny: Exactly, which is something that humans don’t often do, right? You ask somebody to do something for you, they say, “Okay, I got it.” Then they go and give it back to you, and then you have to do the work of being like, “Okay, where did we go wrong? Because what I asked for isn’t what I got.” Where often, like, you said you’re right, even if you’re just talking in text, in some ways, they’re actually very good therapists, because they restate what you said in their own words to you know, convey their act of listening, and then they take action on what they ask for. So it makes it much easier for you to recognize where the miscommunication occurred and sort of try again and see what happens. And that’s a big part of this, right? We live in a world where having the answer is really important, a world where being the first to know how to do something is really important, or doing it, quote-unquote, right is really important. I think 99% of the learning that I’ve had with how to work with this technology is the joy, and I mean that I’m not even being funny, the joy of getting it wrong and figuring it out and then getting it right. And I think that friction that we keep trying to get rid of is actually kind of important, and I worry a little bit what happens in a world where we’re reducing friction but not pushing our own abilities.

Drew: As I’m hearing you talk and thinking about this thing, if we as leaders, think about the amount of times we told someone, “Here’s what I’d like you to do,” and how many times it came back a different way, you realize we’re not very good at communicating to humans. Why would we expect we’d be good at communicating to machines? So the difference is, with the machine, you get a response right away and you go, “Oh, that’s not what I meant.” Whereas with a human they might go off in three weeks and it’s like, “Oh no, you didn’t go the right way.” So the machines will help us communicate, become better communicators, which is not something that I necessarily thought could happen. Let’s get to Split The Decision something you created for a brand because I know endless curiosity you did for yourself, but let’s talk Split The Decision and how that came to be.

Jenny: Yeah, this one was really exciting. I did it for a brand called Zola, which is a wedding planning platform, an excellent company, brand, and product, by the way. They do this survey every year of everybody who sort of joined the platform got married in the previous year. It’s called the First Look Report. And one of the things they found in the First Look Report is that by far the biggest surprise that people had during the wedding planning process was just the sheer volume of decisions that had to be made. There are just way more decisions that had to be made than anybody anticipated. And the biggest sort of pain point issue that people wish was different is that, especially in heterosexual, cisgender relationships, the bulk of that decision-making falls to one person in the partnership, and we can all probably assume who that person is. And I did a lot of research to try to understand what was going on. I spent a lot of time on Reddit and like reading the wedding planning forums, reading the forums where guys don’t talk about wedding planning, because it’s really interesting. Reddit is like 75% dudes, something like that. In most of the subreddits, except for wedding planning, where they’re like 20%, and even in the engagement ring subreddit, it’s mostly women. I thought that’s kind of interesting. But anyways, you know, you see in a lot of these that these women who are incredibly frustrated, incredibly lonely. You see these men who are kind of at sea. They want to help, but they think they’re not supposed to because they think that weddings are a quote-unquote woman’s thing. And we wanted to really do something about this. And so we said, “Well, why don’t we actually make a tool that helps” and so Split The Decisions. What happens is, it’s a custom GPT that’s designed to be done with your partner. You sit down and ask both your names. It uses your names throughout. It asks you when you’re getting married, where are you getting married, about how many people you think might be there. Then it asks each of you the same few questions. Pick three emotions that you have about wedding planning. What are you most excited about? What are you most concerned about? What’s your biggest priority and what do you want your guests to say when they’re leaving the wedding? And ask both of you that, and then it goes through, and it asks a series of which of you questions. So which of you is most likely to actually enjoy negotiating with a vendor? Which of you really cares about things like fonts and colors and aesthetics. And what it does, then is based on all of your answers, on everything. We have this sort of big file of wedding planning tasks, and it basically, sort of divides the tasks equally between you based on the answers that you gave. And it also gives you a downloadable CSV that says the name of the task, the person it’s assigned to. So here you see me and my imaginary fiancé, Julia, and then a link to a Zola tool resource or article to help you pull off the thing that is on that task. And what I think is really cool about this is it basically means that you’re both splitting the duties equally. You’re both doing things that you’re actually the person it’s assigned to is the person most well suited to do it. And it’s a way for this brand to really kind of show up and prove out its purpose with a product that’s in addition to an extension of an offering beyond their core product. And one of the things that I think about a lot with this is, you know, I joke that, after 20 years in advertising, my specialty was always the ideas in the back of a pitch that are really cool and everybody loves but they never get made, and the TV commercials end up getting made, which I think I understood why. Because, you know, when you’re buying media, right? It’s like, okay, I’m going to spend X number of dollars I know as a result, Y number of people are going to see my ad, right? I know that there’s a given with things like this that haven’t been done before that are a little bit different, that don’t have, sort of, they don’t have a built-in guarantee that they’re going to work. And then, on the other side, something like this before could take, you know, months. It could take 10 people, it could take six figures to pull off. And so I think it’s, you know, all my career, it’s been very challenging to do the kind of breakthrough, innovative work that I believe in, because the risk-to-reward ratio is hard, because it’s kind of expensive. And also, I have no way to say, “Yes, I know for a fact that this is going to…”

Drew: There’s several things we want to break down here, and I’m going to break it down first. So you did your homework. You got to understand the customer, and you found a pain point. The bride and the groom are not necessarily equally applied, and the groom isn’t necessarily—might be willing, but doesn’t know where to go. You solved a major problem. You make it easy, and it’s just kind of cool because it’s connected with this brand now. Now, if we step back and we said, huh, insights, those are available, we just have to do our homework. Applications that enable our customers, say buyer enablement, there’s a whole opportunity. And what’s so interesting is, and what Jenny was saying is, I sort of feel your pain, because, you know, running Renegade, a lot of what we got hired to do was the stuff that you were talking about. It was after the ad. But now you can make these things. Now you can create the tools. So there’s one that you can do it, and two, you can do it so cost-effectively and quickly that you can find out whether or not it will work right. It’ll take. So the question, sort of in this is, so far, all of the examples have been GPTs. Was this a GPT as well?

Jenny: Yeah, so this was a GPT for a couple of really specific reasons. You know, we had first started out thinking that this was going to be kind of a standalone product, but once you—as I’m sure everybody on this call knows—try to put something into the platform, it’s like you have the organization. You have to deal with the product team, the IT team, the site team, this team, that team, in addition to the fact that it costs a lot of money and needs a lot of people to develop it. What we kind of came to realize is that, okay, well, if we do this as a custom GPT, it essentially costs us nothing but a $ 20-a-month ChatGPT subscription to be able to create the GPT. And now, one of the things that’s really cool is as of the beginning of this month, custom GPTs can be used by anybody in the world for free. So you basically have free hosting, free media, free development costs for all of that stuff. The reason I do so many custom GPTs is I made something that I think is maybe the next one, or the one after this, and people really liked it. It did really well, and it cost me $1,000 in one week, just on API costs alone, because people were using it, which is a problem to have. But then after that, I was like, “Well, I’m just doing these things as custom GPTs anymore, because I don’t want to have to pay 1000 bucks a week every time we make something that people like.” I’ll stick with my 20 bucks a month, even if it means I have a little bit less control over the kind of visual world of it.

Drew: I think that’s important. If we think about these things as proof of concept, right? So you can develop them quickly. You can create them as GPTs based on an insight. The idea is crystal clear, right? This is a very simple idea. Split the decision, right? Split the work. But it’s brilliant because it is so simple and it is a real problem. This is humans figuring out where the problem was, right? Because you then had to go through all of the exercise of understanding, what are the steps and where are the hiccups, and how do we assign it? It can’t be just, “Hey, you do this one and I’ll do that one.” It can. But this is sort of what do you like to do? So there was a lot of nuanced human-driven insight. You sort of supplemented that, I suppose, by asking GPT for its opinions on these things.

Jenny: Of course I did. Yeah. I mean, that’s, I think, a really good point. One of the beautiful things about this technology is that it is essentially a giant knowledge graph of any recorded human expression or knowledge that the people building these models can get their hands on, right? It’s a huge mind map of the human collective, which is super freaking exciting. At the same time, it’s always, you know, as of right now, a backward-looking thing, right? It’s made of everything that has come before, and so, you know, if you ask it to do the work for you, anybody who’s ever tried to use any of these models to do anything knows that they love VR, AR, they love flash mobs. How funny. I just saw something the other day. I just saw an article from a thought leader in social media who said, “Bad news, everyone, the flash mob is back.” And I laughed because I did wonder how much of that had to do with the fact that these AI language models are always recommending flash mobs, because to them, that’s the pinnacle of advertising, because when they look back, that’s like the most recent sort of viral phenomenon that made it through the training data. And so I always tell people when they’re working with these things that you’re the secret ingredient, your knowledge, your experience, your perspective, your desires. Because these models are, they’re like playing—I don’t know if anybody’s a tennis player, right? It’s like, I could play tennis by myself, right? I could throw the ball, I could hit it, I could run over, I could hit it back. I could run over, I could hit it back. But like, A, that would be exhausting. And B, I probably wouldn’t make many volleys, but I really think for myself of using this technology more as like playing tennis on a backboard, right? I know what I want. I know what I’m trying to get. It’s just a lot faster and more efficient if I can bounce my thoughts off the human collective than me sit alone looking at a wall. But they are, to a certain extent, a mirror, which is a really great thing, and if you know how to use them, that lets you get done more of what you want to do, and bring your ideas to life. But if you don’t have ideas, then all the generative AI in the world isn’t going to help you.

Drew: And the insight to drive the idea, which I think is really important. Again, it wasn’t just, oh randomly, here’s the thing, insight to idea. 

AD Break: This show is brought to you by CMO Huddles, the only marketing community dedicated to B2B greatness, and that donates 1% of revenue to the Global Penguin Society. Why? Well, it turns out that B2B CMOs and penguins have a lot in common. Both are highly curious and remarkable problem solvers. Both prevail in harsh environments by working together with peers, and both are remarkably mediagenic. And just as a group of penguins is called a Huddle, our community of over 300 B2B marketing leaders huddle together to gain confidence, colleagues, and coverage. If you’re a B2B CMO, why not dive into CMO Huddles by registering for our free starter program on CMOhuddles.com? Hope to see you in a Huddle soon.

Drew: Now, if you were building this in GPTs, and as you did in this case, is there any way to track interactions?

Jenny: There are. You can track the number of chats. You could probably put in some sort of custom tracking links, that kind of thing. There are ways you can kind of manage to see what goes from there. Also, OpenAI just acquired an analytics company, so I suspect that’s something that they’re going to be working on, trying to make sure that it’s easier to understand.

Drew: Now let’s move on to the one that is a little more complicated, where you did have an API. I think that’s Plant Advice.

Jenny: This is a custom GPT that uses the chat. The vision had just come out in ChatGPT, where you can upload a picture and it can see what it is. And I have a lot of plants, I’m not always good at taking care of them. And I thought, well, I bet ChatGPT can help me figure out how to take better care of my plants. But I was like, what I really need is a little more motivation to take better care of my plants. And so this gives you advice for your plants. Tells you what your plant needs, but it tells you in the voice of your plant, and the plants are very dramatic. So this one, Dieffenbachia, says, “Do you hear my leaves rustling a melancholic symphony? It is the sound of longing, a plea for the quenching waters that once cascaded through my soil, for the tender touch that would remove the dust from my weary visage. The sun which once danced playfully upon my leaves, now scorches with a relentless gaze. But despair not, dear caretaker, for within me still lies the resilient spirit of the tropics. With attentive care, I can be rejuvenated. Hydrate my parched roots, shield me from the burning sun, and prune the memories of neglect from my form.” And so for me, I’m like, I want to know how my plant’s doing. I feel like shit because I know it’s not doing well, or else I wouldn’t have taken a picture to see how it was doing. So why not make something that gives me the information? Alright, I need to clean off the dead leaves, I need to give it some water, and I need to put it in a more shaded location, but also gives me a smile and gives me a little bit more motivation to do better next time.

Drew: So if you think about this for a moment as a basic Q&A that you have on your website, and you say, “Hey, this is my problem.” And if there was a situation where you can upload an image, but you could put a situation, and rather than it just coming back with, well, you just need to do X, Y, and Z, it came back with a personality and some fun. Imagine how much higher your engagement would be. And so even though this is incredibly silly, and it’s, you know, personal, as in for your plants, you can think about, oh, wow, that’s kind of cool. And the flavor of this, so exactly, how did you get it to create this sort of personality that you have from your plants, so to speak?

Jenny: I told it that the plants were melodramatic guilt trippers. I said I want to hear from the plant. Make sure the plant is a melodramatic guilt tripper.

Drew: How did you get there? Just because you’re an incredibly creative person. But the guilt tripper part just cracks me up.

Jenny: Well, that’s what I wanted. My idea was, what if it could tell you how your plant is doing, but also make you feel guilty by being melodramatic and guilt-tripping? That’s the thing. Like the model will never have that idea without me helping it. But that’s the beauty. I don’t need the model to give me an idea. I have ideas. I have so many ideas, so many ideas. And now that’s the beauty. It used to be that I would have an idea, nobody would be as excited about my idea as me, right? But if I wanted to make it, because I’m not a developer, I’m not an art director, I’m a writer, I’d have to beg and cajole and like, get everybody on board to also think my idea was cool. And, like, a lot of times, I couldn’t get there. Sometimes it would take five years, and by the time I finally got the idea made, it was, like, too late, or I hated it. And now the gap between “I have an idea”—like, I woke up one morning and thought that would be really funny—and two hours later I had this, and it’s something that I can use every day.

Drew: And I just want to make the point for folks is there’s been sort of, I’m not saying that there aren’t a lot of geeks that are creative, but a lot of the assignment of using these tools has been for the folks that may have come from engineering. They may be coming from some aspect of it because it feels like technology. But I think that you also need to make sure that your most creative team members or you have external partners, are bringing that both insight and creativity to the use of the tools, because it’s garbage in, garbage out. If you can’t bring creativity at the top end, you’re not going to get it in the back end. Okay, let’s keep moving. Examples. Storyteller.

Jenny: This was a fun one. ChatGPT, they released actions, which are the ability to sort of basically let you make API calls and stuff like that. If anybody’s familiar with the API, what’s called function calls, if you’re using the OpenAI API, are actions in ChatGPT, and I wanted to play with them, so I made this tool called Spotify Storyteller, where you give it a title of a story, and it writes a story where the story is actually a Spotify playlist. So this one, I think the title I gave it was “The vegetables in your fridge are mad because you didn’t eat them and they’re going bad.” And so then it actually created this playlist, wrote it to Spotify, added the songs, gave it a description just for me, saying “The vegetables in your fridge are mad about going uneaten. The unhappy vegetables: a narrative playlist capturing the journey of vegetables as they go from being fresh to their eventual decay, told from their perspective.” So we have “Fresh in the Dark,” “Decaying,” “Silo V2,” “Wasted Summers,” “Throw It Away,” “Crying Alone.” “Now this was kind of a fun one, because, like, did I need to make this? No, but I wanted to experiment with actions. I wanted to say, “Okay, I want to see what I can make.” I was like, what would be entertaining to me? And then I have this idea. And I was like, “Alright, let’s try it.” And it was also sort of an interesting lesson in the pros and cons and pluses and minuses, because this one was one of the ones that took me a while, because OpenAI, God bless them, sometimes their implementations are a little sloppy, sometimes their documentation is a bit lacking. And so it took me a long time to get it hooked up, and I finally got it working after, like, a week and a half, I was so excited. And then I discovered that the Spotify API Terms of Service, they have a dev mode which limits you to like five people, but you have to invite them by email. And if you want to get it to more people than that, you have to go through this application thing, and then they also in this fine print said, “And you can’t actually do anything that uses generative AI in any way, shape or form.” So I found out after I had already built this, that the API Terms of Service meant that this was something I was never going to be able to put out into the world, but it was an awesome learning experience. It was an awesome challenge. I’m really pleased with that, and I think it’s probably only a matter of time before they end up shifting those terms of service.

Drew: Just remind the audience, you are not a coder, right?

Jenny: No, I did teach myself way, way back in the day, in like 2000, how to make a website in like Notepad. But I’m not a developer. I just did it because I was curious and wanted to know if I could. I guess the best way to describe myself, and it feels incredibly egotistical to use a quote from Albert Einstein, but I’m going to because it’s the most true thing I’ve ever seen. And he said, “I have no special talents. I am only passionately curious.” And that’s the best description of the way I approach the world that I could come up with. I did it because I wanted to see if I could. And for me, a lot of times that’s enough.

Drew: I think this show needs more boom.

Jenny: Okay, so this is the one that cost me $1,000 per week and made it so that I stopped making websites. So this is a website. It’s called needsmoreboom.com and also I can share this with you later, Drew. All of these that have the underline, I actually made a link for people to try it.

Drew: Awesome. Thank you.

Jenny: So this you basically put in any movie scene that you can think of, and it gives you a script where the scene is rewritten as though it were a Michael Bay movie. So you know, one of my favorites is the scene in Lion King where Mufasa dies, and what happens is, like, Mufasa does still die, but usually, like, Scar comes in wearing a leather jacket and aviator glasses riding on like a mechanical wildebeest that’s equipped with a rocket launcher. But what’s really fun about this, too, is it doesn’t just take movie scenes, like I did one, where it was the instructions for assembling an Ikea bookcase, and it was this ridiculous scene where it ends with a giant explosion, and the bookcase lands on the ground completely assembled as a result of the explosion. Somebody did like the Microsoft end user licensing agreement terms, and it just wrote this whole scene of somebody trying to understand it and losing their mind. So it was just kind of a really fun example of how to use this technology. What I think is cool is this website. I built it by myself. I used a tool called Bubble, which let me do the design, make all the things, set it up so that when you put in your thing, it got passed to the API that got the result, and the result got passed back, and it got sent in front of you, like a lot of things that, you know, maybe 10 years ago, I would have been too intimidated to even try. I was able, I think this whole thing took about 10 hours that people could actually use, that they could go to, that they could play with.

Drew: 10 hours to build a single website that had this application “Needs More Boom?” 

Jenny: Yes

Drew: We all need more boom, but the credit that I have for the audience is what aspect of your business or your customers needs more boom? What are these fun ways of engaging, whether it’s your employees or your community of customers, or your partners, where a little bit of boom? And again, I’m being facetious with the boom, but it’s not about the simplicity of the idea and the way that you can have a little fun in the middle of your day.

Jenny: And I’ll give you an example, right? So all of these companies are rolling out, like, large language model-powered customer service chatbots, right? So many of these chatbots, like, pretend to be human, and I’m like, You know what? Like, I know it’s likely AI. You know it’s likely AI. Anybody using these things, even if they are talking to a human these days, assume that they’re talking to an AI, pretending to be a human. And I’m like, why not just let it be AI and lean into the best of it? Like, I wish that I could go to a website and I could decide what kind of customer service. Is my customer service chatbot going to be friendly and warm? Is it going to be like just the facts ma’am, or, you know, is it going to be an alien who meets all of my needs but continually farts and awkwardly apologizes for it? The great thing about that is like they all give you the same information. They all follow the same rules. They’re all using the same sort of FAQ to inform their responses, but like, I get to have some sort of say over the experience that I have. So if I want a response that’s just like, robotic and straightforward, if I want somebody to pretend to be a human, friendly person to me, or if I want something that feels like a joke, I still get the information, but I have a little bit more control over the experience, and maybe it feels a little bit more enjoyable. All because we didn’t try to make AI act like it was a human. We said, “Okay, if this is an AI customer service chatbot, what could we do with it that we would not have been able to do if it was just a human on the other end of the chat.”

Drew: Choose your flavor, choose your adventure. One of your sort of instincts is, okay, let’s turn it into a game. So talk about that one.

Jenny: I’ll turn anything into a game. This is called TaskFlare. It’s also a custom GPT, and what it does is you just put in everything you have to do in one day, and it automatically reorders your list from the easiest thing to the hardest thing, and then for each item on the list, it generates an image of that task as a metaphorical monster that then you get to slay it by saying I did the task. So here I think I had to take out the trash, so that’s our red-eye glowing trash monster. And here I had to fold the laundry. So here’s the scary monster coming out of the dryer. And why not? Like, getting things done is hard and it sucks, and I don’t want to do it. And anything that I can do that makes it a little bit more fun, that makes it a little bit more engaging, that makes it a little bit more interesting. Like, why? Why wouldn’t we? I did something for an agency, everybody had to do their yearly OKRs. And if anybody’s been in an organization where people have to do OKRs, you know that people don’t like to do them, which is interesting, because OKRs is an employee for you, how are you going to ask for a raise if you don’t have an agreed-upon thing that you’re working toward and agreed-upon result that says that you reached it? And so they said, “Can we make something that gets people more excited about or more engaged or more willing to do their OKRs?” And so we ended up making a chatbot called The Art of OKRs. And what it does is it already knows all of the agency-level OKRs. It knows what the agency goals are, but it comes in and it basically asks you to be really selfish. It says, What are your really selfish goals? What do you want to achieve? Is it a raise? Is it a promotion? Is it to work on this or that? It helps you figure out what you selfishly want, and then it helps you tie your selfish desires to the agency’s goals so that you have OKRs that are meaningful to you but also relevant to what the agency is trying to do. But the funnest part is it does the whole thing in the tone and voice of Sun Tzu and Machiavelli.

Drew: What a combination.

Jenny: They’re very good advisors.

Drew: From very different points of view. What is interesting to me about this is there’s a moment where you onboard a customer and you have an opportunity to sort of either get them involved in using the product to its fullest potential, or you have an opportunity to sort of lose them, because there’s a lot of potential buyer’s remorse at the moment that they start to use it and so forth. So something like this that works as an educational program that helps people get through it to any way that can make some of the experiences of learning a new software, say, for example, or the steps that you have to take to take advantage of something or just creating to-do lists in general, and then priorities. So there’s a lot of different ways that you can go with this application. And thinking about it again, I’m really thinking and focusing on employee experience, employee education, customer experience. There’s a whole other world of buyer enablement in helping them. But a good place to start and play, I think, is in your internal audiences. Is there one more sort of favorite creation that you have that we should cover?

Jenny: Project Runway GPT edition is one of my favorites.

Drew: Okay, talk about that.

Jenny: So this is another vision, one where you upload a picture of your outfit, and then the judges of Project Runway tell you if your outfit’s good or not, and they give it an in or an out, and if they vote out, they tell it how to make it an in. And also, I’m very pleased with this AI-generated Heidi Klum cyborg with the actual spelled-right Project Runway logo. So I wanted a little credit for that. I could just look in the mirror. I could just like, ask GPT. I could just not worry about it. But this idea that I can take something that I’m a big fan of and make something that is both useful and functional, but also really fun and engaging. I think that’s what I would love to see more of, because honestly, like our world these days, is filled with a lot of boring stuff that’s not fun, and if we could find a way to turn even the sort of the most boring moments of onboarding or task completion or whatever into something that actually feels like a fun and engaging experience, then I think that’s a huge, a huge world of opportunity that I wish more, more companies were paying attention to.

Drew: Yeah, I mean, I’ve been thinking about so you’re rebranding your company. Employee Education is really hard. Folks are doing their job and they don’t necessarily pay attention, but if they don’t get inculcated, if they don’t embrace the new brand, it will often fail. So some companies deploy tests, and I love that in principle, but it’s kind of harsh, whereas this could be a sort of a more fun version of helping educate employees on the purpose.

Jenny: Yep, one of the very, very first games I ever made for work was for the liquor brand Chambord, because we had been asked to help them with their, I think, was the distributors, the guys who drive around to the restaurants and the stores and the bars like getting them to stock up Chambord. And they asked us to do a training program for them. And I think they were like a PowerPoint. I was thinking and I was like, man, like, I don’t know if anybody knows anything about Chambord. It’s this raspberry liqueur that comes in this, like, kind of royal-looking purple globe with, like gold gilt and, like a little crown top. It’s a kind of kind of lady-ish. And I was like, man, what a job these guys just like, drive around in their cars all day, like, getting people in like bars at like one in the afternoon to buy their like liqueur. I’m like, man, that seems hard. I was like, what if we made them cool and we ended up making an interactive game called Operation Chambord, where they operate as, like a double-oh-seven style agent, and they get a little cool car to drive around in, and they get to change the music on their car’s radio. And like, they go in, and there’s a module in the game where, like, they have to watch a video on the history of the Chambord estate in the Loire Valley in France, and some lady, kind of comes in and brings a bottle of Chambord to them, and you pick it up in the game. And the bottle of Chambord is a secret video player, and it plays the video on that. Or there’s a scene where you have to figure out where to put the shelf talkers in the liquor store, and you do that, but before you do that, you have to kill three ninjas with your space bar. And so it’s just how you take something that has to be done and actually give it a little bit of something that respects the time of the person who’s doing it, I think goes a long way. And I remember the completion I think they have, like, a 98% completion rate of the game, and people said it was the best thing they’d ever done at work. And I think about that a lot. This was back in like, I don’t know, maybe 2006-2007 and we could have just given them a PowerPoint, and they probably would have been happy with that, but then we would have missed an opportunity to do something so much cooler.

Drew: So the moral of the story is, we’re going to use the machines to create more human experiences. Hmm, that’s a really fascinating takeaway. Speaking of takeaways, let’s end with two do’s and one don’t when it comes to using generative AI. And let’s put a narrow time frame on it for B2B brands.

Jenny: Two do’s. One do is experiment. Start small. You know, if you think about sort of effort to impact, right? Companies right now are going all the way, like full effort, going all in, transforming our whole system, and then they’re paralyzed, because it’s like, how do you do that on a technology moving so fast, something new comes out every week. How do you transform your entire organizational pipeline? You don’t, you can’t it’s too much. It’s too much. And you don’t know where to put your bets right now, because everything is changing, right? A week ago to a month ago, it would have been really safe to put your bet with OpenAI, but over the last month, Anthropic has been moving really fast, and so your bet might want to change there. And so experimenting in small ways is better than trying to go all in because it’s not an all-in time yet. We don’t know yet. The other thing I would say is this technology, you know, when you’re doing something in an organization, it’s very tempting to sort of have your task force get your pipelines and products in place, buy something, then roll it out down through your organization. But this technology actually works the opposite, right? I really strongly believe that the biggest transformations aren’t going to happen in companies that get new pipelines or processes. They’re going to come to companies that give everybody on their team access to this technology and understanding on how to use it, and then a charge to figure out what they can do with it, because I always say, I do a lot of these trainings, and I’m like, it’s not like learning Photoshop, it’s not like learning Salesforce. It’s like learning how to use a pencil, right? I can teach you what a pencil is. I can teach you what it’s made of. I can show you how to sharpen it. I can show you how the eraser works. I can show you what happens if you press down too light, or if you press down too hard, but at the end of the day, what you’re going to do with that pencil is different than what I’m going to do with it. And I think that’s the thing that people are missing about this technology, is that instead of trying to make something that you force everybody into what if you give people access to it, and if every single person who uses it finds a way to make themselves 20% more productive. What does that do for the organization as a whole? And what does that do for the morale and the excitement of everybody who works there when they’re being rewarded, encouraged, you know, spotlit and plotted for what they’re doing and what they’re discovering, versus when there comes in, and there’s this top-down thing that everybody has to use. And, oh, by the way, like you just push a button and a machine does all the work. How does that go for morale? And the one thing I think you can’t do is ignore it. I think that’s the one thing you can’t do.

Drew: So one of the things I would just add to this, and so we’re going to be smaller, quicker experiments. We are going to try to disseminate it out to the groups. But you know, you mentioned being more get 20% more productive. I want 20% more creative too.

Jenny: Maybe the person who’s using it isn’t a creative person, right? And not everybody is a weirdo like me, who has 10 ideas before they wake up, but they might have 10 ideas on how to do something smarter and faster, it’s when everybody is being forced to try the same thing in the same way that the real benefit of this time in human history is going to be lost.

Drew: There it is. Alright. Jenny Nicholson, how can people find you and engage with you?

Jenny: You can find me on LinkedIn, but then you can also find me at queenofswords.co

Drew: Alright, and how do they engage with you?

Jenny: I do a couple of things. Primarily, one of the first ways companies engage with me is by training. So I do virtual sessions, I do team workshops, I do one-on-one training, and then I also do building so say you have an idea, or you don’t have an idea, and you want to come in and actually make something, experiment with something, build something kind of like what I did with Zola, like I can actually build it for you. So that’s the thing that I think is really cool about what I did with Zola, is that was about a three-week project, and it was just me and one person on the client side at Zola, and the two of us built that whole thing together, and we got a lot of good PR from it. We got a lot of coverage. We got a lot of good feedback, and now my client at Zola is actually building a business case to make Split the Decisions an actual official integrated part of their product offering.

Drew: Awesome. Well, I want to thank you, Jenny, and I want to thank everyone here in the audience for staying with us and listening, and hopefully you are leaving just a little bit more inspired about the opportunities ahead.

If you’re a B2B CMO and you want to hear more conversations like this one, find out if you qualify to join our community of sharing, caring, and daring CMOs at CMOhuddles.com.

Show Credits

Renegade Marketers Unite is written and directed by Drew Neisser. Hey, that’s me! This show is produced by Melissa Caffrey, Laura Parkyn, and Ishar Cuevas. The music is by the amazing Burns Twins and the intro Voice Over is Linda Cornelius. To find the transcripts of all episodes, suggest future guests, or learn more about B2B branding, CMO Huddles, or my CMO coaching service, check out renegade.com. I’m your host, Drew Neisser. And until next time, keep those Renegade thinking caps on and strong!